Previous research, including studies emerging from collaboration between the principal investigator and co-principal investigator, has suggested that a syntax-semantics mapping deficit may underlie the agrammatic language production disorder found in non-fluent Broca's aphasics. A form of treatment for agrammatism known as mapping therapy has emerged which targets the link between grammatical functions such as subject or object and thematic roles such as Agent or Theme, and significant treatment gains have already been reported. A large-scale study investigating the efficacy of mapping therapy is currently underway in the co-principal investigator's laboratory. The Unisys-Moss collaboration proposed here builds directly upon this work, with the following objectives: 1) To develop a computer-based mapping therapy system for agrammatism which targets the full range of psycholinguistic operations involved in this mapping: provides immediate, linguistically explicit feedback about the accuracy of patients' responses; and systematically manipulates both animations and sentence materials to reinforce mapping based upon grammatical structure rather than extralinguistic strategies; 2) to test the efficacy of this system through single case studies of ten agrammatic patients, tracking both intra-linguistic generalization (treatment effects from one linguistic structure to another) and extra- linguistic generalization ('scaling up of treatment effects to increasingly challenging context); 3) to analyze patterns of intra- linguistic generalization with particular concern for evidence about whether generalization occurs on the basis of phrase structure or predicate argument structure shared by the trained and untrained structures; 4) to develop and incrementally incorporate into this therapy system the capability for the automatic recognition and interpretation of agrammatic speech. The Unisys spoken language understanding system PUNDIT, of which the principal investigator is a developer, will serve as the basis for this research.

Agency
National Institute of Health (NIH)
Institute
National Institute on Deafness and Other Communication Disorders (NIDCD)
Type
Research Project (R01)
Project #
1R01DC002377-01A1
Application #
2127696
Study Section
Sensory Disorders and Language Study Section (CMS)
Project Start
1995-06-01
Project End
1998-05-31
Budget Start
1995-06-01
Budget End
1996-05-31
Support Year
1
Fiscal Year
1995
Total Cost
Indirect Cost
Name
Unisys
Department
Type
DUNS #
City
Paoli
State
PA
Country
United States
Zip Code
19301
Linebarger, M C; Schwartz, M F; Romania, J R et al. (2000) Grammatical encoding in aphasia: evidence from a ""processing prosthesis"". Brain Lang 75:416-27